Optimal design of wheel profile for railway vehicles

Abstract

Beyond the advantages in design of running gear and railway vehicles itself and introducing the active wheelset steering control systems many railroads and tram companies still use huge number of old fashion vehicles. Dynamic performance, safety and maintenance cost of which strongly depend on the wheelset dynamics and particularly on how good is design of wheel and rail profiles. The paper presents a procedure for design of a wheel profile based on geometrical wheel/rail (w/r) contact characteristics which uses numerical optimization technique. The procedure has been developed by Railway Engineering Group in Delft University of Technology. The optimality criteria formulated using the requirements to railway track and wheelset, are related to stability of wheelset, cost efficiency of design and minimum wear of wheels and rails. The shape of a wheel profile has been varied during optimization. A new wheel profile is obtained for given target rolling radii difference function ' r y ∆ − ' and rail profile. Measurements of new and worn wheel and rail profiles has been used to define the target ' r y ∆ − ' curve. Finally dynamic simulations of vehicle with obtained wheel profile have been performed in ADAMS/Rail program package in order to control w/r wear and safety requirements. The proposed procedure has been applied to design of wheel profile for trams. Numerical results are presented and discussed.

An Innovative and Efficient Method for Reverse Design of Wheel-Rail Profiles

“…However, this method is very time-consuming. Shevtsov et al [11,12] optimized the design of wheel profiles by taking the rolling radius difference as the design target according to multipoint approximation technology based on response surface fitting (MARS) and improved the profile matching and dynamic performance. Reference [13], based on the target functions and optimization method [11,12], optimized railway wheel profiles considering the wheel-rail rolling contact fatigue and wear etc.…”

“…Shevtsov et al [11,12] optimized the design of wheel profiles by taking the rolling radius difference as the design target according to multipoint approximation technology based on response surface fitting (MARS) and improved the profile matching and dynamic performance. Reference [13], based on the target functions and optimization method [11,12], optimized railway wheel profiles considering the wheel-rail rolling contact fatigue and wear etc. Jahed et al [14] proposed a similar method to that used by Shevtsov et al [11,12] and they chose a reduced set of generalized coordinates constituted by five variable parameters together with cubic spline curves in a linear programming formulation for the generation of profiles.…”

“…Reference [13], based on the target functions and optimization method [11,12], optimized railway wheel profiles considering the wheel-rail rolling contact fatigue and wear etc. Jahed et al [14] proposed a similar method to that used by Shevtsov et al [11,12] and they chose a reduced set of generalized coordinates constituted by five variable parameters together with cubic spline curves in a linear programming formulation for the generation of profiles. The objective is to minimize the deviation of the rolling radii difference of the generated profiles from the target one.…”

An inverse shape design method for railway wheel profiles

“…To solve the minimisation problem, an optimisation method that uses multipoint approximations based on response surface fitting (MARS) method was used. In the studies of Shevtsov et al (2003Shevtsov et al ( , 2005, the procedure was further developed, and three approaches for choosing the target RRD function were suggested. Similar approach was proposed in Shen et al (2003) wherein the contact angle function was used for design of railway wheel profiles.…”

“…In this paper, a procedure described in the studies of Shevtsov et al (2003Shevtsov et al ( , 2005) is applied to improve the performance of metro vehicles that were suffering from severe wheel wear and instability (Markine et al 2004). The RRD function used in the formulation of the optimality criteria is described in the Section 2.…”